Influence of Drainage on Peat Organic Matter: Implications for Development, Stability, and Transformation

The agricultural use of peatlands, the stabilization of the substrate for building or road construction or for increasing the capacity of soil to support heavy machinery for industrial activities (peat and petroleum extraction), harvesting to provide peat for energy, and the growing media and initiation of chemical processes must be preceded by drainage. As a consequence of drainage, peat underwent an irreversible conversion into moorsh (secondary transformation of the peat). The object of the study was to investigate comparatively the organic matter composition and molecular structure of humic acids (HAs) in the raised bog, fen, and peat-moorsh soils developed in various compositions of botanical cover, peat-forming species, and oxic and anoxic conditions as a result of the oscillation of ground water during drainage as well as to evaluate the vulnerability of soil organic matter (SOM) to decomposition. Drainage was shown to be the principal factor causing the various chemical compositions and physicochemical properties of HAs. Large and significant differences in chemical composition of peat and the properties of HAs were found to be related to the degree of decomposition. The HAs from drained peatlands were less chemically mature. In contrast, the HAs from fen and raised bog were found to be more mature than that of the corresponding drained peatlands. The above findings showed the distinguishable structure of HAs within the soil profile created by the plant residue biodegradation and formed in both oxic and anoxic conditions. The analytical methods of thermal analysis together with the optical densities and paramagnetic behaviour are suitable and effective tools for studying structure–property relationships characterizing the origin and formation process of HAs in various environmental conditions.


Introduction
There are peatlands in every climatic zone. Total Earth's area of these elements of the landscape is more than 3 million km 2 (about 3% of the Earth's land surface). Peatlands accumulate some 70% of natural freshwater. Each year, they absorb 0.37 giga-tonnes of carbon dioxide (CO 2 ), which corresponds to a greater capacity for carbon storage than that attributed to all other vegetative organisms in the world [1].
The conversion of vegetable matter into peat structures changes depending on the vegetation type, ability of tissues to be decompose, deepness of roots and water table, pH, ionic strength, balance making them more susceptible to water-table fluctuations and droughts. A further aspect of fen habitat degradation is acidification. This may be related to drainage, which results in partial replacement of groundwater by rainwater [50], and to the increased atmospheric deposition of nitrogen and sulphur compounds [51].
In general, drainage of peat leads to the progressive differentiation of the hydrophobic peptides and total amino acid content in organic matter. In proteins of peats, hydrophobic contacts exist between hydrophobic and hydrophilic structural elements (between the side chains of the radicals of phenylalanine, leucine, isoleucine, valine, proline, methionine, and tryptophan). Hydrophobic forces stabilize the tertiary structure of proteins and determine the properties of lipids and biological membranes. The presence of amino acids, hydrocarbon chains, and other nonpolar fragments in their composition are related to hydrophobic properties of humic substances [37,52].
Since organic matter is a major component of the soil phase of peat and moorsh soil causing soil water repellency, it is important to study the effect of chemical soil properties on their wettability. It was observed that the significant changes of chemical properties of transformed organic matter in peat have a significant influence on the sequential modifications in physical and hydraulic features created by lowering of water table for agriculture. Van Dijk [53] postulated relationship between a high increase of shrinkage, changes in the number of many chemical and physical properties and humic components in peats [53].
The retention of water by peat can be considered in terms of reaction of water molecules with the surfaces of peat particles. The phenomenon is therefore amenable to the analytical methods of colloid and surface chemistry. Considerable attention should be paid to the chemical composition of peat, identifying the molecular structure of substances and aggregates most likely to hold water strongly. Knowledge of this structure can provide a rationale for treatments intended to remove or render less water-retentive the most hydrophilic fraction.
The substances of greatest immediate interest are humic acids (HAs). HAs are created in peats by degradation, poly-condensation, polymerization, and poly-addition of organic substances as a result of habitat and anthropogenic processes, including the degradation of plants and animal residues which are characterized by a complex macromolecular structure with aromatic and aliphatic units; peptide chain; and nitrogen in aliphatic, cyclic, and aromatic forms. The HAs represent macromolecular polydisperse biphyllic systems, including both hydrophobic domains (saturated hydrocarbon chains and aromatic structural units) and hydrophilic functional groups, i.e., having an amphiphilic character. The hydrophilicity of peat surfaces is generally attributed to the availability of organic functional groups capable of hydrogen-bonding. Such groups, well known to organic chemists, include carboxyls as well as phenolic and alcoholic hydroxyls. HAs have been regarded as peat component principally responsible for water retention in peat. The hydrophilicity of HAs depend not only on the numbers of hydroxyl and other polar units but on their ability to form hydrogen-bonding with water as there are hydroxyl groups inaccessible to water.
These substances are formed of similar but not identical substrates; therefore, no two HAs are identical. The HAs from various types of peats and organic and inorganic components of HAs matrices are altered to different extents, significantly differing both in composition and properties. The quantity and quality of HAs in soils organic matter depend on the balance between primary productivity and the rate of decomposition [38]. Chemical properties and structural characteristics of humic substances were shown to be better predictors of soil organic matter turnover rate in vertisols than soil organic matter content.
This suggests the possibility of using humic substances as indicators of soil organic matter turnover because they are sources of intermediates and energy for many chemical and biochemical pathways in the soil [39,40].
The study outlined here was conducted to propose tools and analytical methods for the quantitative and qualitative evaluation of the turnover processes occurring in the oxic and anoxic conditions of developed peat deposits. The aim of this study is to analyse comparatively the organic matter composition and molecular structure of HAs in the raised bog, fen and peat-moorsh soils developed in various composition of botanical cover, peat-forming species, and oxic and anoxic conditions as a result of the oscillation of ground water during drainage as well as to evaluate the vulnerability of soil organic matter (SOM) to decomposition.

Study Sites
Six sites from 3 peatlands were found to vary according to their macrofossil analysis, their state of decomposition, and their GPS parameters, and an outline of their location has been given ( Figures 1 and 2, Table 1). The study outlined here was conducted to propose tools and analytical methods for the quantitative and qualitative evaluation of the turnover processes occurring in the oxic and anoxic conditions of developed peat deposits.
The object of the study was to comparatively investigate the organic matter composition and molecular structure of HAs in the raised bog, fen, and peat-moorsh soils developed in various compositions of botanical cover, peat-forming species, and oxic and anoxic conditions as a result of the oscillation of ground water during drainage as well as to evaluate the vulnerability of soil organic matter (SOM) to decomposition.

Study Sites
Six sites from 3 peatlands were found to vary according to their macrofossil analysis, their state of decomposition, and their GPS parameters, and an outline of their location has been given (Figures 1 and  2, Table 1).   The study outlined here was conducted to propose tools and analytical methods for the quantitative and qualitative evaluation of the turnover processes occurring in the oxic and anoxic conditions of developed peat deposits.
The object of the study was to comparatively investigate the organic matter composition and molecular structure of HAs in the raised bog, fen, and peat-moorsh soils developed in various compositions of botanical cover, peat-forming species, and oxic and anoxic conditions as a result of the oscillation of ground water during drainage as well as to evaluate the vulnerability of soil organic matter (SOM) to decomposition.

Study Sites
Six sites from 3 peatlands were found to vary according to their macrofossil analysis, their state of decomposition, and their GPS parameters, and an outline of their location has been given (Figures 1 and  2, Table 1).

Collection of Smples (WRB classification 2015)
Peat samples were collected from the following: a) Baltic-type raised bog (Kusowo) b) fen (Stążka) c) peat-moorsh soils: Ch1, Ch2, Ch3, and Ch4 (Turew) Peat samples were collected in triplicate from the field using a 5.0-cm diameter Instorf peat auger and various depths from 0 to 100 cm in the stratigraphic profile of each peat deposit, transported to the laboratory at ca. 4 • C and stored at −20 • C. The samples were dried at 20 • C and homogenized in a grinder after removal of any visible live plant material, after which they were passed through a 1-mm sieve to remove rock fragments and large organic debris. The botanical composition of peat was analyzed by the microscopic method and subsequently classified according to the Polish standards (PN-76/G-02501 1977). Peat samples were used for the description, classification, and physicochemical analysis [54] ( Table 1).
The Kusowo Bog is situated in the West Pomeranian Voivodship. This is likely the best-preserved Polish Baltic-type raised bog. The reserve has an area of 326.56 ha and is entirely included into the Szczecineckie Lake Natura 2000. The mean peat thickness may exceed 12 m. The mean annual air temperature is 7.2 • C. The mean annual precipitation is 760.1 mm. (Table 2) ( Figure 1). The age of peatland is 710 years ( Table 3). Moisture contents of peat samples ranged from 89.69% to 92.31% (Table 5). Among raised bogs, that of the Baltic-type is distinguished, one peculiar to a humid climate, with a high rainfall. The bog, formerly a lake, developed on a moraine with kames in an extensive depression. The bog is clearly divided into two parts-northern and southern-separated by three mineral mounds. The highest average annual precipitation was measured for Baltic-type raised bog (Kusowo) at 760.1 mm, and the lowest was measured for peat-moorsh soils at 539.3 mm. plant association Sphagnetummagellanici. In places with high water content, e.g., near peat ponds, the Rhynchosporetumalbae, Eriophoroangustifolii-Sphagnetumrecurvi, and Caricetumlimosae are located. Nearly half of the northern part consists of wooded habitats, primarily pine and birch bog forests (Vacciniouliginosi-Pinetum and Vacciniouliginosi-Betuletumpubescentis) [55,56] (Figure 1) The Stążka fen is a part of the "Bagna nad Stążką"-a mire complex located in Northern Poland in the region of the Tuchola Forest. The fen is part of the Nature Reserve, where the whole complex of natural peatlands is under protection. This peatland measures some 478.45 ha. The maximum thickness of peat deposits is 1.4 m. This region was characterized by a mean annual air temperature of 7.2 • C and by mean annual precipitation 598.9 mm (Table 2) [26]. The age of peatland is 1400 years ( Table 4). Moisture contents of peat samples ranged from 93.15% to 93.92% (Table 5).  Drainage significantly impacted the concentration of the moisture in peats. The moisture contents in undrained peats (90-94%) were higher than in drained peats (61-83%) ( Table 5).
Peat-moorsh samples were taken from four chosen sites marked as Ch1, Ch2, Ch3, and Ch4 on the 4.5-km long transect of peatland located in the Chłapowski Agro-ecological Landscape Park of the West Polish Lowland, about 40 km southwest of Poznań. These are the sites investigated along the Wyskoć Canal. The thickness of peat deposit ranges from 1.5 to 2.75 m. The mean annual air temperature is 8.4 • C, the mean annual precipitation is 539.3 mm (Table 2), and the growing season is about 200 days (Figures 1 and 2). Moisture content of peat samples ranged from 61% to 86.82% (Table 5).
All examined samples were classified as raised bog, fen, and peat-moorsh soils according to World Reference Base Soil Resources (2015) [57]. Air-dried peat samples for the measurements of hot water extractable organic carbon (C HWE ) were mixed with deionized water and heated at 100 • C for two hours under a reflux condenser. Extracts were filtered through 0.45-µm pore-size filters and analyzed on TOC 5050A facilities, Shimadzu, Japan [58].
Total nitrogen was evaluated by the Kjeldahl methods, using Vapodest 10s analyser (Gerhardt, Germany). N-total was evaluated by semimicro-Kjeldahl method; 0.5 g of soil sample was set in Kjeldahl digestion tube. Five mL of twice distilled water was added. After 30 min, 2.5 mL of 95% H 2 SO 4 and powder zinc were added and heated 20 min. Fifteen mL of 95% H 2 SO 4 , 5 g of K 2 SO 4 , and a small amount of selenium mixture were added and heated at 350 • C. A Kjeldahl digestion tube was cooled to room temperature. Mineralized sample was set in a 100-mL volumetric flask and filled to the line by twice distilled water. A mineralized sample was centrifuged (MLW K 23 D, Germany) in 3000 rpm by 15 min. Twenty-five mL of the sample was set in Parnas-Wagner apparatus. Fifty mL of 30% NaOH was added. Fifty mL of distillate was collected to a 100-mL Erlenmeyer flask filled by 20 mL of 4% H 3 BO 3 and indicator (bromocresol green and methyl red mixture). The blue distillate was titrated to red color by 0.02 N HCl.
The atomic ratios of C/N were calculated after determining the concentrations of carbon and nitrogen.
Peat decomposition was determined at the sampled depths according to the field squeezing method by means of the von Post classification scale [59]. This method identifies ten classes of decomposition, with H1 being undecomposed peat and H10 being completely decomposed peat. Peat type was determined based on plant macrofossil analysis [60].

Extraction and Purification of HAs
Isolation of HAs was performed according to the recommendations of the International Humic Substances Society procedure under relatively gentle conditions [61,62]. The humic substances were extracted from air-dried peat soils using the grain size fraction smaller than one millimeter with 0.1 M NaOH using an extractant/peat ratio of v/v 5:1) at pH 7.00 under a N 2 gas atmosphere and shaken for four hours and, after that, were stored overnight for the coagulation of HA fractions which were separated by centrifugation. The purification of the HAs was performed by the following method. The suspension was centrifuged at 4000× g at 24 • C for 1 h. The solution was acidified with 6 M HCl to pH 1.3 to precipitate the HAs and was allowed to stand overnight. Then, the solution was centrifuged to eliminate the supernatant. The procedure was repeated three times. Finally, precipitated HAs were freeze dried and stored in a vacuum desiccator over P 2 O 5. Moisture analyzer MAX series (Radwag, Poland) was used to determine hygroscopic humidity content in peats. The content of hygroscopic humidity was taken into account in the analysis of peats.

Elemental Composition
The content of C, N, H, and S in every HAs fraction was analysed using the Vario Micro Cube Elemental Analyser. The samples were burned in an oxygen atmosphere at 1150 • C. The gases (CO 2 , N 2 , H 2 O, and SO 2 ) were separated chromatographically and measured by a thermal conductivity detector (TCD). Oxygen content was obtained by subtracting the sum of other elemental contents from 100%, and the results were expressed as percentages calculated on the basis of total element amounts ( Table 6).

Electron Paramagnetic Resonance of HAs
Electron paramagnetic resonance (EPR) X-band (9.8 GHz) was recorded at room temperature on a Bruker ElexSys E500 instrument equipped with an NMR teslameter ER 036TM and E 41 FC frequency counter. The concentration of spins was measured using a double integral procedure applying the Bruker WinEPR program. Leonardite was used as a spin concentration standard (International Humic Substance Society) [64].

Differential Thermal Analysis of HAs
Thermal properties of HAs were evaluated by means of an OD-103 derivatograph (MOM-Paulik-Paulik-Erdey, Hungary) [65]. The curves of differential thermal analysis (DTA), thermogravimetry (TGA), and differential thermogravimetry (DTG) were recorded simultaneously. Weight losses at different steps of thermal decomposition were calculated from the TGA curves. The Z index was calculated, and aliphatic character of HAs was pointed out. All chemical analyses were run in triplicate, and the results were averaged.

Statistical Analysis
The confidence intervals were calculated using the following formula: x ± t α (n − 1) SE, where: x is the mean; t α (n − 1) is the value of the Student test for α = 0.05; n−1 is the degree of freedom; and SE is standard error. Linear correlations between the values were calculated. Normal distribution of the results and homogenous variances were checked before statistical analysis.
Principal component analysis (PCA) using Statistica version 9.1 was performed to determine the correlations between Baltic-type raised bog, fen, and peat-moorsh soils in the peat deposit properties and physicochemical properties of soil organic matter. The number of factors extracted from the variables was determined by a scree test according to Kaiser's rule. With this criterion, the first two principal components with an eigenvalue greater than a third were retained. Principal component analysis (PCA) using Statistica version 9.1 was performed to determine the correlations between raised bog, fen, and peat-moorsh soils and of peat deposits properties and physicochemical properties of soil organic matter.

Decomposition Degree, Carbon, Nitrogen, and C/N Ratio of Peat Soils Decomposition Degree
Degree of decomposition is an important property of the organic matter in soils and other deposits which contain fossil carbon. It describes the intensity of transformation or the humification degree of the original living organic matter. Our knowledge of the degree of decomposition and chemical characteristics of peat soil may be translated into peat physicochemical properties, therefore, both the degree of decomposition and type of peats described above [7,66].
All peats were decomposed, however, in varying degrees (Table 1). A gradual increase of the degree of decomposition has been observed in the order from the Baltic-type raised bog (H2-H5) throughout fen (H3-H5) to peat-moorsh soils (H7-H8). A drop of the water level in the peat-moorsh soils increased the oxygen content and accelerated the mineralization of organic matter, leading to an increase of the degree of decomposition (H8). Therefore, a decrease in the degree of peat decomposition is treated as an indicator of a rise of the ground water table [67].
During decomposition of organic matter, labile compounds such as peptides and polysaccharides are preferentially decomposed while refractory aromatic or aliphatic compounds become residually enriched [68].
Almost all nitrogen in organic matter is in organic form (0-90%). Nevertheless, the chemical composition of nitrogen in organic soil fraction is not completely understood and little is known of the factors affecting the distribution of organic nitrogen forms in soils.
Humus is composed from 20% to 60% of HAs. The nitrogen (20-40%) in HAs consists of amino acids or peptides, the main unit of proteins, and is connected to the central core by hydrogen bonds [69]. Therefore, amino acids and proteins are the main fraction of total nitrogen in organic soils.
Soil amino acids are components of protein conglomerates. They occur in stable form. Soil proteins included in organic colloids are hydrophilic colloids. These colloids are water-related. Drainage causes the denaturation of colloids and a change in the properties of proteins from hydrophilic to hydrophobic. During the drainage process, a progressive increase in the hydrophobic amino acid content was observed. In proteins of peats, hydrophobic contacts exist between hydrophobic and hydrophilic structural elements (between the side chains of the radicals of phenylalanine, leucine, isoleucine, valine, proline, methionine, and tryptophan). The presence of amino acids, hydrocarbon chains, and other nonpolar fragments in their composition are related to hydrophobic properties of humic substances [37,52].

TOC
Significantly increasing concentrations of TOC exhibited a trend with a gradually increasing down-core for all investigated sites (Table 1). A lower TOC was determined in peat-moorsh soils in comparison with Baltic-type raised bog and fen. This relationship was found to be related to the degree of decomposition. These changes observed for the degree of decomposition, and TOC could have been caused by a change in any one of a number of the following factors, e.g., water level and the balance between accumulation/decomposition, indicating most likely that they are attributable to the combination of such factors.
The lower TOC content in the upper layer may be related to the increased rate of decomposition rather than accumulation. In deeper layers of peat profile under saturated conditions, lower decomposition rates of organic matter are observed.
Our results are in line with the results of Benavides [70], who showed that peat and carbon accumulation rates were lower in drained sites, indicating either greater decomposition rates of the upper peat column or lower production by the changed plant communities. The ecological services offered by peatlands to agrarian communities downstream are important. In addition, he observed that species composition was much affected by drainage, which resulted in a reduction in cover of Sphagnum and other peat-forming species, and by the encroachment of sedges and Juncus effusus. The ability of peat to store water and carbon was also reduced in drained peatlands. Vegetation records show a shift towards sedge-Juncus communities around 50 years ago when agricultural use of water increased.
Batjes [71], Chambers et al. [72], Fontaine et al. [73], and Wang et al. [74] suggested that a lack of a supply of fresh carbon may prevent the decomposition of organic carbon pool in deeper layers of peat deposits. C HWE C HWE is responsible for the microbiological activity in depth profile to be exuded from plant roots, acting as a significant source of carbon fueling microbial metabolism [75]. This fraction is a potential source of carbon and energy for heterotrophic organisms and contributes significantly to stream ecosystem metabolism. The flux of C HWE from an ecosystem can be a significant component of carbon (C) budgets, especially in watersheds containing wetlands. In watersheds containing organic wetland soils or peatlands, the flux from the watershed can be 4-8% of annual net primary production, a significant fraction that should be addressed when performing a carbon mass balance [76].
A significant decline of C HWE with increasing depth in all sites was observed ( Table 1). The highest decrease of C HWE (68.2%) with increasing depth was measured in Ch2, representing peat-moorsh soil, and can be attributable to the most degraded peat-moorsh soil.
In contrast, the lowest decrease of C HWE (25.4%) with increasing depth was determined in the lowest decomposed peats of Baltic-type raised bog as compared to that of peat-moorsh soils. Thus, C HWE can be said to increase gradually with the increase of peat during decomposition ( Table 1).
The decreases in C HWE concentrations were generally accompanied by the increases of the TOC in all sites of sampling. We found a consistent pattern in parallel changes as per decreasing/increasing C HWE and TOC concentrations among all sites.
The values of the ratios C HWE: /TOC in undrained peat (raised bog and fen) ranged from 0.9 to 2.7, while in drained peat, it ranged from 1.2 to 4.1. The highest value of the ratio, 4.1, was measured in the Ch2, which showed the highest value of the degree of decomposition.
Monitoring the properties of C HWE in the peat profile is frequently used to evaluate changes in peat quality and to explain shifts in peatland ecosystem function [77].
The purification of ground water by the transect consisting of peat-moorsh soils of 4.5 km length was observed (Figure 2). Peatland plays a positive function as a biogeochemical barrier, which reduces the content of chemical compounds moving in the groundwater throughout the peatland. In groundwater between Ch3 and Ch2, the concentrations of nitrates (38.5%), N-organic (10%), N-total (24.5%), ammonium (38.7%), dissolved total carbon (33.1%), dissolved total inorganic carbon (10%), and dissolved organic carbon (57.5%) were significantly decreased [78]. C/N Another effective and common indicator of peat decomposition is the use of C/N ratios [79]. This indicator is based on the observed residual enrichment of N relative to C during mineralization of organic matter.
The concentrations of N-total in Baltic-type raised bog were significantly lower compared to the N-total in fen and in peat-moorsh soils (Table 1). Baltic-type raised bog receives water and nutrients primarily from atmospheric deposition.
Significant differences of C/N ratios between less decomposed Baltic-type raised bog, fen, and highly decomposed peat-moorsh soils were observed ( Table 1). The C/N values in undrained peats were significantly higher as compared to that of corresponding drained peats. The highest C/N ratio and the lowest degree of decomposition in the Baltic-type raised bog was measured.
The high C/N ratios in the deeper layer of peat indicates more intensity in the accumulation of organic matter with the formation of mature HAs as compared to that in the upper layer. In contrast, the low values of the C/N in the surface layer of all peat-moorsh soils correspond to drier conditions and the intensity of organic nitrogen mineralization, leading to the formation of gaseous substances such as ammonia and the further emission of N 2 O and N 2 into the atmosphere [80][81][82].
Baltic raised bog and fen represent two types peats with similar values of the degree of decomposition H2-H5 (Table 1). Baltic raised bog as nitrogen deficiency system was shown to have lower concentrations of N total and higher C/N ratios compared with that of the corresponding fen. Drainage was shown to be the principal factor causing the increase of the degree of decomposition and the decrease of C/N ratios. Peat-moorsh soils were created as a result of secondary transformation of fen. Therefore, in peat-moorsh soils, higher degrees of decomposition H7-H8, higher C/N ratios, and lower concentrations of TOC compared with that of the corresponding in fen were measured. For all peat-moorsh soils, the values of decomposition degree were shown to be in line with the results calculated for C/N ratios (Table 1).
Our results are in line with the results of Malmer and Holm [83] and of Kuhry and Vitt [84]. They pointed out that lower C/N ratios characterize more decomposed peat material.
According to Broder et al. [85], the lowest humification and high C/N ratios in Sphagnum bogs justified the high polyphenol content. On the other hand, Rice and MacCarthy [86] and Anderson [81] pointed out that the C/N ratio does not always reflect soil organic matter mineralization and, moreover, that the percentage share of C WHE would seem to be a better indicator of organic matter mineralization as compared to the C/N ratio (expanding the trend in the C/N ratios with depth), which is bound to hydrological changes of nitrogen concentration.

Elemental Composition of HAs
One of the most fundamental characteristics of HAs is its elemental composition expressed in the weight and atomic percentage structure of particular elements [86,87].
The C content ranged from 41.79% to 47.04%, from 40.76% to 44.38%, and from 36.09% to 46.03% in Baltic-type raised bog, fen, and peat-moorsh soils, respectively ( Table 6). The C concentration increased with an increase in the depth profile for highly decomposed peat, while the opposite was shown for oxygen content, which decreased in all peat profiles with increasing depth.
The H and N amounts here demonstrated higher variability. The S concentration decreased with increasing depth in Baltic-type raised bog and fen. In contrast, the concentrations of S increased with increasing depth in peat-moorsh soil compared to the increasing depth in Baltic-type raised bog and fen.
The ratio of H/C, C/N* (in HAs), and O/C bears much more diagnostic information than the elemental composition of HAs. The magnitude of the H/C ratio has been used to indicate the degree of aromaticity (a small value) or aliphaticity (a large value) of a substance [52]. The values of H/C ratio in studied HA samples are in the range from 1.20 to 1.28 in Baltic-type raised bog, from 1.12 to 1.24 in fen, and from 1.00 to 1.25 in peat-moorsh soils (Table 6). However, these H/C and C/N* values among peat soils are not significantly different.
The decrease of the H/C ratio in HAs with increasing depth in undrained peats (Baltic-type raised bog and fen) is related to the peat accumulation rate and development of the aromatic structures, where more labile structures are destroyed or transformed into the appearance of more stable aromatic and polyaromatic structures.
On the other hand, in drained peat (peat-moorsh soil) with increasing depth, the increase of the H/C in HAs increased, indicating less amounts of the C in HAs as compared to that of undrained peats (Baltic-type raised bog and fen).
The C/N* and O/C ratios increased with increasing depth. The high values of C/N* in the deeper layer are related to the presence of proteinaceous materials of living organic matter, indicating a higher intensity of humification than mineralization process with the formation of mature HAs ( Table 6).
The significant negative correlations between the O/C vs. C/N* (r = −0.448) and H/C vs C/N* ratio (r = −0.632) indicate that, with an increase in depth, the decarboxylation processes were in line with the increase of the N concentration in HAs (Table 9).
According to Van Krevelen [88], the H/C atomic ratios from 0.7 to 1.5 correspond to aromatic systems coupled with aliphatic chains and contain up to ten carbon atoms. The O/C ratio, for its part, is considered as an indicator of carbohydrate and carboxylic group contents and can be directly related to the aromatization of the peat-forming organic matter. DiDonato et al. [87] suggested that more of the carboxyl-containing aliphatic molecules are sourced from lignin.

VIS spectra of HAs
The E 4 /E 6 ratio is a valid and informative index for the characterization of aromatic condensation and poly-conjugation in the humic molecule [35,89,90]. Table 7 shows that the E 4 /E 6 ratio of HAs fraction in peat-moorsh soils ranged from 5.18 to 6.95; in fens, from 4.46 to 6.2; and in Baltic-type raised bog, from 3.98 to 6.36. In peat-moorsh soil profiles, the variability of this value was generally low. As expected, the E 4 /E 6 ratio of HAs decreased with increasing depth profile in all investigated sites related to progressive humification and condensation of aromatic constituents. These findings are in line with our data of elemental analysis of HAs, indicating more chemically mature HAs in the bottom rather than in the upper layers.
Moreover, the E 4 /E 6 trend was found to be higher in drained peats compared to E 4 /E 6 from undrained peats. This reflects a lower degree of aromatic condensation and poly-conjugation and a lower degree of humification in the molecules of HAs from drained rather than undrained peats [88,91,92].
In this context, Kl , avin , š and Sire [93] showed strong negative correlations between total acidity values and the E 4 /E 6 ratios in peat bog profiles. They pointed out that an increase of the acidic groups in the HAs samples resulted in a reduced E 4 /E 6 ratio.

Electron paramagnetic resonance of HAs
Our study shows the impact of peat type and the degree of decomposition on the EPR spectra, which consists of one signal with ∆H pp at about 4 G typical for radical species (Figure 3) (Table 7). g-value, the spin concentrations are given in spins/g units, e.g., the last value is 3.41 × 1017 spins/g organic matter.
Moreover, the E4/E6 trend was found to be higher in drained peats compared to E4/E6 from undrained peats. This reflects a lower degree of aromatic condensation and poly-conjugation and a lower degree of humification in the molecules of HAs from drained rather than undrained peats [88,91,92].
In this context, Kļaviņš and Sire [93] showed strong negative correlations between total acidity values and the E4/E6 ratios in peat bog profiles. They pointed out that an increase of the acidic groups in the HAs samples resulted in a reduced E4/E6 ratio.

Electron paramagnetic resonance of HAs
Our study shows the impact of peat type and the degree of decomposition on the EPR spectra, which consists of one signal with Hpp at about 4 G typical for radical species (Figure 3) (Table 7).  A strong aromatization in the deeper layer is distinctly observed for the Baltic-type raised bog. The drop of oxidative conditions in the depth has led to an increase in EPR signal intensity for HAs ( Table 7) and decline of the g-value from 2.0035 to 2.0021.
The maximum of spin concentrations determined in the deepest layers are in line with the VISspectra of HAs and reflects more chemically mature HAs from the Baltic-type raised bog compared with the fen and peat-moorsh soils.
Our results are in line with Czechowski and Jezierski [94], who observed a steady decrease in A strong aromatization in the deeper layer is distinctly observed for the Baltic-type raised bog. The drop of oxidative conditions in the depth has led to an increase in EPR signal intensity for HAs ( Table 7) and decline of the g-value from 2.0035 to 2.0021.
The maximum of spin concentrations determined in the deepest layers are in line with the VIS-spectra of HAs and reflects more chemically mature HAs from the Baltic-type raised bog compared with the fen and peat-moorsh soils.
Our results are in line with Czechowski and Jezierski [94], who observed a steady decrease in the g parameter towards the g = 2.0023 value of free electrons for free radicals in bituminous carbon components in parallel with an increase in the aromaticity of the free radicals. In addition, the effect of g parameter reduction (determined by EPR studies) of semiquinone radicals naturally occurring in HAs of increasing aromaticity (Knüpling at al. [95]) was confirmed and explained theoretically by Witwicki and Jezierska [96] on the basis of quantum mechanical calculations of g parameters for model semiquinone radicals.
Inspection of EPR spectra of HAs from fen and peat-moorsh soils (Ch1 and Ch3) reveals g-value characteristics for HAs (2.0036) [64]. The deeper layers of these deposits include an increasing concentration of semiquinone radicals (Table 7) connected with the formation of oxygen containing polymers without growing aromaticity, as the g-values are practically unchanged. In addition, the same g-values of the HAs in all deeper layers are related to similar redox conditions.
The sample Ch2 (peat-moorsh soil) reveals a rather unexpected change of the spin concentration that appeared to be highest in the upper layer. The oxidative conditions of the upper layer impacted the g-value and the spin concentration in this HAs. The g-value is smaller for the upper layer (2.0035) as compared to the g-values (the difference exceeds the experimental error) for lower layers (2.0036), showing a slightly greater aromaticity. This phenomenon is related to the high degree of decomposition of the upper layer; the most transformed peat soil from all studied peat-moorsh soils; the highest O/C ratio in molecule of HAs; and the highest TOC, C HWE , and N-total (Table 1). In the context of CH2, due to a high in situ, the oxidative properties in the decomposition rate of the organic matter are faster than its accumulation.

Thermal Properties of HAs
All shapes of DTA and DTG curves are compatible. Thus, each thermal effect on the DTA curve is related to a weight loss measured in TGA of thermal degradation of HAs exhibiting endothermic and exothermic peaks associated with these in all samples ( Figure 4; Table 8) [97][98][99]. The DTA curves of all HAs showed one endothermic effect and three exothermic effects. Inspection of EPR spectra of HAs from fen and peat-moorsh soils (Ch1 and Ch3) reveals gvalue characteristics for HAs (2.0036) [64]. The deeper layers of these deposits include an increasing concentration of semiquinone radicals (Table 7) connected with the formation of oxygen containing polymers without growing aromaticity, as the g-values are practically unchanged. In addition, the same g-values of the HAs in all deeper layers are related to similar redox conditions.
The sample Ch2 (peat-moorsh soil) reveals a rather unexpected change of the spin concentration that appeared to be highest in the upper layer. The oxidative conditions of the upper layer impacted the g-value and the spin concentration in this HAs. The g-value is smaller for the upper layer (2.0035) as compared to the g-values (the difference exceeds the experimental error) for lower layers (2.0036), showing a slightly greater aromaticity. This phenomenon is related to the high degree of decomposition of the upper layer; the most transformed peat soil from all studied peat-moorsh soils; the highest O/C ratio in molecule of HAs; and the highest TOC, CHWE, and Ntotal (Table 1). In the context of CH2, due to a high in situ, the oxidative properties in the decomposition rate of the organic matter are faster than its accumulation.

Thermal Properties of HAs
All shapes of DTA and DTG curves are compatible. Thus, each thermal effect on the DTA curve is related to a weight loss measured in TGA of thermal degradation of HAs exhibiting endothermic and exothermic peaks associated with these in all samples ( Figure 4; Table 8) [97][98][99]. The DTA curves of all HAs showed one endothermic effect and three exothermic effects.   DTA-differential thermal analysis; DTG-differential thermogravimetry; Z-the loss ratio of HAs in the low range to those in the high temperature range; endo-endotermic; exo-exotermic effect.

of 27
The endothermic effect ranges from 74 to 84 • C, with weight loss from 12% to 20% and corresponding to the evaporation of absorbed water and dehydration reaction. This effect does not demonstrate any significant changes downsizing into the peat profile, and the intensity of weight losses is related to weakly bounded water by hygroscopic components.
Our results are in line with Schnitzer et al. [100], Kodama and Schnitzer [101], Shurygina et al. [102], Leinweber and Schulten [103], and Gołębiowska et al. [65], who pointed out that, at low temperature (about 100 • C), the evaporation of hydroscopic moisture is observed. In addition, Kl , avin , š and Sire [59] showed that the amount of hygroscopically bound water in HAs from peat is higher compared to peat from the corresponding layer and does not differ in peat profile.
The thermal analysis of HAs in all peat samples showed an increasing combustion temperature with a downsizing depth of the profile for three exothermic effects (Table 8), which are related to more thermally labile fractions in HAs of the upper layer compared to bottom layers.
The first exothermic reaction ranged from 277 to 320 • C with a mass loss from 27.4% to 31.6% related to the thermal combustion of carbohydrates, peptides, lignin, external functional groups of HAs, decarboxylation of acidic groups in aliphatic compounds, and dehydration of hydroxylate aliphatic structures.
The second exothermic reaction varied from 348 to 427 • C with a mass loss from 9.1% to 32.4%, corresponding to the combustion of less mature components of HAs.
The third exothermic reaction ranged from 437 to 496 • C with a mass loss from 24.6% to 41.4%, which is related to the combustion of bitumens, to highly mature HAs of increasing thermal stability, and to the cleavage of C-C bonds [104,105].
In contrast, in Ch1 and Ch2 (peat-moorsh soils with the highest degree of decomposition), the mass loss of the second exothermic effect increased with an increase in depth.
The lowest mass loss was measured for the second and the highest mass losses for the third exothermic effect in Ch2 with the highest degree of decomposition.
Schnitzer and Levesque. [66] detected two exothermic peaks between 200 and 500 • C for humic substances of podzolic soil. The authors hypothesized that the first peak was associated with dehydratation and decarboxylation and that the second was associated with carbon oxidation.
The thermal analysis provides some important information on HAs structure based on Z parameter values, which expresses the ratio between thermally labile and stable parts in HAs (mainly aliphatic and aromatic) [97].
Inspection of Z parameters revealed that, with increasing depth, HAs contain more structures resistant to oxidation in high temperatures. The HAs from Ch2 and Ch3 (peat-moorsh soils) revealed a more aliphatic character in contrast to the CH1 and Ch4 Baltic-type raised bog and fen, which were found to be more aromatic ( Table 8).
The significant negative correlation between the heat of the combustion (DTA 1 /DTG 1 ) and E 4 /E 6 (r = −0.528) ( Table 9) may be attributable to more condensed stable aromatic structures of HAs in the deeper layers of peat deposits.
The thermal analysis of HAs in all peat samples showed an increasing combustion temperature with a downsizing depth of the profile for three exothermic effects (Table 8), which are related to more thermally labile fractions in HAs of the upper layer compared to bottom layers.
The first exothermic reaction ranged from 277 to 320 • C with a mass loss from 27.4% to 31.6% related to the thermal combustion of carbohydrates, peptides, lignin, external functional groups of HAs, decarboxylation of acidic groups in aliphatic compounds, and dehydration of hydroxylate aliphatic structures.
The second exothermic reaction varied from 348 to 427 • C with a mass loss from 9.1% to 32.4%, corresponding to the combustion of less mature components of HAs.
The third exothermic reaction ranged from 437 to 496 • C with a mass loss from 24.6% to 41.4%, which is related to the combustion of bitumens, to highly mature HAs of increasing thermal stability, and to the cleavage of C-C bonds [104,105]. In contrast, in Ch1 and Ch2 (peat-moorsh soils with the highest degree of decomposition), the mass loss of the second exothermic effect increased with an increase in depth.
The lowest mass loss was measured for the second and the highest mass loss for the third exothermic effect in Ch2 with the highest degree of decomposition.
Schnitzer et al. [100] detected two exothermic peaks between 200 and 500 • C for humic substances of podzolic soil. The authors hypothesized that the first peak was associated with dehydratation and decarboxylation and that the second was associated with carbon oxidation.
The thermal analysis provides some important information on HAs structure based on Z parameter values, which expresses the ratio between thermally labile and stable parts in HAs (mainly aliphatic and aromatic) [97].
Inspection of Z parameters revealed that, with increasing depth, HAs contain more structures resistant to oxidation in high temperatures. The HAs from Ch2 and Ch3 (peat-moors soils) revealed a more aliphatic character in contrast to the CH1 and Ch4 Baltic-type raised bog and fen, which were found to be more aromatic ( Table 8).
The significant negative correlation between the heat of the combustion (DTA 1 /DTG 1 ) and E 4 /E 6 (r = −0.528) ( Table 9) may be attributable to more condensed stable aromatic structures of HAs in the deeper layers of peat deposits.
The findings illustrated in Table 9 show a significant negative correlation between TOC concentrations and Z values (r = −0.645), which is related to the increasing contents of aromatic structures in line with the downsizing of peat profile. The above data agrees with E 4 /E 6 and C/N values (Table 9), where the decrease of both ratios with an increase of depth corresponded to a high degree of aromatic condensation and poly-conjugation of aromatic structures and to the presence of relatively low proportions of aliphatic structures.
As shown in the study of thermal organic matter stability [106], these properties are a function of the chemical composition, degree of humification, and mineral association. However, Kl , avin , š and Sire [93] showed that the decay of more labile structure in HAs decreases as the degree of peat decomposition increases, though more condensed aromatic structures were positively correlated with the peat degree of decomposition.
Francioso et al. [107] showed in peat HAs various biochemical fractions of plants preserved during peat formation characteristic of thermal decomposition and more labile structures in the origin and formation process of HAs. This preservation therefore might be the result of anoxic environmental conditions occurring due to peat accumulation.

PCA Analysis
A variety of multiple causes are responsible for the composition of peat and the properties of HAs from undrained and drained peats. Therefore, a comparative study was carried out of the chemical composition of peats and the properties of HAs from various undrained and drained types of peats developed in various compositions of botanical cover, peat-forming species, and oxic and anoxic conditions as a result of the oscillation of ground water during drainage.
The PCA analysis was conducted to determine the impact of various types of peat and decomposition degree on the properties of peat deposits and HAs. The results reveal that the properties of organic matter and HAs vary in different environments ( Figure 5).
The results of the PCA explain 80.26% of the total variability of organic matter properties. The first two axes were statistically significant (α < 0.05): the first axis (PC1) explains 45.98%, and the second (PC2) explains 19.84% of the total variability, while the third axis (PC3) explained 14.43% (Table 10). The PC1 was negatively associated with the TOC, C/N, spin concentration DTA 1 /DTG 1 , and C/N* and positively coordinates with E 4 /E 6 , g-value, and carbon in HAs on this axis in peat soils. The PC2 showed a close negative association with C HWE and H/C. chemical composition of peats and the properties of HAs from various undrained and drained types of peats developed in various compositions of botanical cover, peat-forming species, and oxic and anoxic conditions as a result of the oscillation of ground water during drainage.
The PCA analysis was conducted to determine the impact of various types of peat and decomposition degree on the properties of peat deposits and HAs. The results reveal that the properties of organic matter and HAs vary in different environments ( Figure 5). The results of the PCA explain 80.26% of the total variability of organic matter properties. The first two axes were statistically significant (α < 0.05): the first axis (PC1) explains 45.98%, and the second (PC2) explains 19.84% of the total variability, while the third axis (PC3) explained 14.43% (Table 10). The PC1 was negatively associated with the TOC, C/N, spin concentration DTA1/DTG1, and C/N* and positively coordinates with E4/E6, g-value, and carbon in HAs on this axis in peat soils. The PC2 showed a close negative association with CHWE and H/C.   The results of the PCA analysis showed on PC1, with 45.98% variability, a significantly positive correlation between TOC, spin, and DTA 1 /DTG 1 ( Figure 5). The data suggest significantly an increased TOC, spin, and DTA 1 /DTG 1 with the depth, indicating a decrease in the content of thermolable structural units (carbohydrates, free, and bound functional groups) and an increase in thermostable skeleton part of the HAs molecules. This reflects that the transformation of the organic matter is strongly connected with the humification process and the molecular structure of HAs. Moreover, there were significant negatively associate of TOC, spin, and DTA 1 /DTG 1 with E 4 /E 6 . The observed changes can be explained by a higher degree of aromatic condensation and poly-conjugation in the molecules of HAs, in the deeper layers (Table 10).
Additionally, on the second axis PC2 with 19.84% variability, the C HWE vs. H/C was significantly positively associated ( Figure 4). The decrease of C HWE content and H/C values with the depth suggests the decline of aliphatic bridges between aromatic structural units HAs. On this basis, it can be stated dominantly the accumulation processes rather than ones of decomposition with the depth of peat profile.
The third axis showed a significant positive correlation with ΣDTA/ΣDTG (Table 11). These variables suggest that the first principal component describes the dependence of VIS-spectra, electron paramagnetic resonance, and thermal properties of HAs on the organic matter. Therefore, a higher g-value and E 4 /E 6 in peat-moorsh soils as well as fens compared to those from Baltic-type raised bogs affects a lower degree of aromatic condensation and humification of HAs.

Conclusions
This study has examined the vulnerability of organic matter of Baltic-type raised bog, fen, and peat-moorsh soils to decomposition by determining the chemical composition and physicochemical properties of HAs.
Drainage was shown to be the principal factor causing the chemical composition and physicochemical properties of HAs. The latter properties of HAs were found to be in line with their chemical composition of peats.
Conversion of undrained to drained peatlands (modifications in oxic-anoxic properties) has led to an increase of the degree of decomposition, relating to the mass loss of organic matter, particularly labile organic fractions, and was confirmed by a correspondingly lower content of TOC, lower C/N values, and higher E 4 /E 6 values. In line with an increase of the degree of organic matter decomposition, the N-total; E 4 /E 6 , g-values; mass loss for exo1, exo 2, and exo 3; Z values; and spin concentrations all increased while C/N decreased.
In light of these results, the HAs from undrained peatlands were found to be richer in aromatic structures and can be regarded therefore as more humified and mature (therefore, more stable) than that of peat-moorsh soils. Overall, the most important indicators for the propensity of organic matter to decomposition as a consequence of drainage identified in the present study were more aliphatic structures in the HAs. There would appear to be therefore a positive feedback loop of drainage and the consequent vulnerability of soil organic matter to decomposition.
In this context, the analytical methods of thermal analysis together with optical densities and paramagnetic behaviour can be said to be suitable and effective tools for characterizing the origin and formation process of HAs in various environmental properties.